Location of Hand Function in the Sensorimotor Cortex: MR and Functional Correlation C. Rumeau, N. Tzourio, N. Murayama, P. Peretti-Viton, 0. Levrier, M. Joliet, B. Mazoyer, and G. Salamon PURPOSE: To determine the location of hand function in the sensorimotor cortex using MR and positron emission tomography imaging studies. METHODS: Anatomic and physiological meth.ods were used for this study. Anatomic study was based on the MR analysis of 22 subjects. The length of the sensorimotor cortex was measured in the axial and sagittal planes. Physiologic study was based on the positron emission tomography studies of 4 subjects. Each of the studies was correlated with MR. RESULTS: We found that the superior genu of the central sulcus corresponds to hand function in the sensorimotor cortex. This level may prove useful for any clinical correlations or for surgery. CONCLUSIONS: From this study, the hand function area in the sensorimotor cortex is easily understood with its characteristic shape in axial MR scan. The comparison of MR and positron emission tomography data clearly show anatomic correlations. This may be applied to the functional mapping of the pathologic studies in the sensorimotor cortex regions. Index terms: Magnetic resonance, experimental; Positron emission tomography (positron emission tomography); Brain, magnetic resonance; Brain, anatomy AJNR Am J Neuroradio/15:567-572, Mar 1994 Magnetic resonance (MR) imaging is one of the small degree of individual variation. This re­ most effective studies used to analyze the detailed stricted definition considers the sensorimotor area configuration of the cerebral cortex. Single-pho­ as limited to the precentral gyrus and postcentral ton emission computed tomography (SPECT) gyrus. The present study analyzes both anatomic and positron emission tomography (PET) scan and functional points of view. The purpose of this techniques, through determination of regional study was to determine the location of hand variations of the cerebral blood flow, allow as­ function in the sensorimotor cortex by correlating sessment of cortical areas activated during the MR and positron emission tomography imaging. performance of specific sensorimotor activities. Anatomically, the sensorimotor area constitutes a rather easily identifiable structure located Materials and Methods around the central sulcus, between the pre- and Anatomic Study postcentral sulci. The precentral gyrus is limited by the precentral sulcus and the central sulcus. MR Acquisitions. Twenty-two healthy adult volunteers (19 to 51 years old) were examined in 0.5-T or 1.5-T MR The postcentral gyrus is limited by the central scanners. All were right-handed and had no history of sulcus and the postcentral sulcus. It displays a neurologic disease. To maintain strictly the orbitomeatal plane during the examination, the head was immobilized against a suitable head support with adhesive tape. MR Received November 17, 1992; accepted pending revision February 8, scanning was performed in two steps. In the first step, the 1993; revision received April 14. reference marks of the anterior and posterior commissures From the Department of Neuroradiology, Hospital La Timone, Mar­ seilles, France (C. R., N. M. , P. P.-V., 0. L., G. S.); and the Service were identified by doing sagittal Tl-weighted sequences Hospitalier Frederic Joliot et Groupe d'lmagerie Neurofonctionnelle, S. H. centered on the midline. In the second step, 19 contiguous F. J ., CEA, Orsay, and Hopital R. Debre, CHU Bichat, Universite Paris VII , sagittal planes with a 7-mm-thick section and 16 contig­ Paris, France (N. T ., M. J. , B. M.) uous axial and coronal planes with a 7-mm-thick section Address reprint requests to Pr G. Salamon, CHU La Timone, Bd Jean were obtained orthogonal to the sagittal plane. The MR Moulin, 13385 Marseille Cedex 5, France. Fax: (33) 91 85 83 17 scans were acquired as inversion recovery (2000/500/2 AJNR 15:567-572, Mar 1994 0195-6108/ 94/ 1503-0567 [repetition time/ echo time/ excitations]), inversion recovery © American Society of Neuroradiology time 21, with an image matrix size 256 X 256 mm. 567 568 RUMEAU AJNR: 15, March 1994 HORIZONTAL PLANE AC M P C SAG I TTAL PLANE 53 . A Fig. 1. Segmentation of the profiles of sensorimotor cortex in the axial plane. The length of each segment was measured in the whole series of transparent drawings. 1 indicates precentral sulcus; 2, superficial part of the precentral gyrus; 3, precentral gyrus (depth of the central sulcus); 4, postcentral gyrus (depth of the central sulcus); 5, superficial part of the postcentral gyrus; and 6, postcentral sulcus. For every MR plane, identification of the sulci was determined by experienced neuroradiologists. All the struc­ tures identified were referenced according to the commis­ sural plane. B Sensorimotor Cortex Morphologic Quantification. As shown in Fig 1, the contours of the left sensorimotor cortex were divided into six segments: 1) precentral sulcus, 2) surface of precentral gyrus, 3) frontal slope of precentral gyrus, 4) parietal slope of postcentral gyrus, 5) surface of postcentral gyrus, and 6) postcentral sulcus. The widths of all the segments included in the whole series of sections for each subject were measured on enlarged pictures in the axial and sagittal planes using a Hewlett Packward 9830 A computer, connected to a Hewlett Packward 9864 A x-Y plotter. The widths obtained were checked against those of a brain directly measured by a curvometer that followed the surface of the brain. The total length of the sensori­ motor cortex was calculated by summing up the lengths c of the six segments included in either axial or sagittal sections. The motor cortex was defined as the sum of Fig. 2. The axis of the central sulcus in axial and sagittal planes segments 1, 2, and 3 because it is necessary to take into in one case. account the deeper part of the motor cortex. Following this A, On the axial plane, the axis forms a frontal facing open angle of 47° with anterior commissure-posterior commissure (AC­ same reasoning, the sensory cortex was defined as the sum PC) . On the sagittal plane, this angle measures 53°, and is open of segments 4, 5, and 6 to take into consideration both the toward the occipital pole. 8, MR on axial plane showing the superficial and the deeper part of the sensory cortex. measurement of the angle. C, Angle measurement on sagittal MR. AJNR: 15, March 1994 HAND FUNCTION 569 the autoattenuation of the 511 kev gamma rays in the subject's head. It is also used to reconstruct "transmission" images that are in fact equivalent to low contrast computed tomography (CT) scans and that were used to match the PET and the MR coordinate systems. Cerebral blood flow experiments consisted of repeated intravenous bolus injections of 50 mCi of oxygen 15-labeled water (1, 2). For each injection, data were acquired in the time span of 2 minutes. For each subject the experimental protocol included six injections with a constant 20-minute delay between two injections. Subjects were kept with their eyes closed and their ears plugged; no order was given except to relax. Two conditions were alternated: "rest" and "vibration" of the fingers using a vibrator operating at 130 Hz (3). Vibration was started at least 30 seconds before the Fig. 3. The location of the opercular region and anterior com­ water-labeled injection and lasted 2 minutes. Subjects were missure-posterior commissure line. 1 indicates frontal operculum; told not to grasp the vibrator, but instead to let the 2, rolandic operculum; 3, parietal operculum; PRCG, precentral experimenter gently maintain their fingerpads in contact gyrus; PSTCG, postcentral gyrus; AC, anterior comissure; and with the vibrator. After each cerebral blood flow acquisition, PC, posterior comissure. data were read from a disk and a single scan (7 axial sections) was reconstructed that started at the arrival of the radioactivity in the brain (identified as the time of a sharp rise in the total number of counts registered by the tomograph) and that lasted for 80 seconds. Emission (cer­ ebral blood flow) and transmission (attenuation) PET im­ ages were reconstructed using the same size , 128 X 128 with a 2-mm-square pixel size. Reconstruction was per­ 1 .. formed using a backprojection algorithm and a 5 mm- _7~~+ j-- cutoff frequency Hanning filter that translates in a 7-mm in-plane resolution at image center. Image processing was performed on Microvax II computers coupled to Ramtek 9465 graphic workstations. Image Analysis. The image analysis procedure that was Fig. 4. Sagittal view of the brain showing the projection of designed to use MR individual studies and to match ana­ sensorimotor area and central sulcus. The lengths of the profiles tomical (MR) and functional (PET) images has been de­ of the unfolded sensorimotor area are also shown to demonstrate scribed elsewhere (4, 5). Briefly, after conversion to the that the unfolded sensorimotor cortex exceeds the anteroposterior same format, PET and MR images were superimposed by length of the brain between +49 and +63 mm from anterior aligning two sets of isodensity contours. commissure-posterior commissure line. First, MR axial sections were converted to PET image size ( 128 X 128, 2 mm-square pixel size). The identification Functional Study of section levels between the two modalities was checked A functional study was done on four healthy right­ using the PET transmission images and the corresponding handed volunteers 21 to 28 years old. The axial MR and MR sections and superimposing isodensity contours cor­ PET images were acquired on a plane parallel to the responding to the skin defined on MR and on transmission orbitomeatal line. The exact position of the head was sections. Adequacy was visually checked and was found systematically controlled during each MR or PET scan satisfactory in all cases.